Elsevier's position paper on Open Access

[Jan Velterop <jan@biomedcentral.com>]

(Cross-posted) 

http://www.elsevier.com/authored_news/corporate/images/UKST1Elsevier_positio
n_paper_on_stm_in_UK.pdf

The link above is to Elsevier's position paper on Open Access. Elsevier is
a fine company, of course, with fantastic profits, but on the issue of
Open Access publishing they are curiously ill-informed. Lest unnecessary
misunderstandings about Open Access or BioMed Central arise, we feel that
a response to their position paper is in order. The page numbers refer to
the Elsevier document.

Page 1 

[Elsevier:]* Access: all UK Higher Education Institutions engaging in
science and medical research and all researchers within them have access
to nearly all Elsevier journals that pertain to their research programmes:
97% of UK researchers have direct access, on average, to around 90% of
Elsevier journals under licence of their host institution. UK citizens
have access to all Elsevier journals and articles either directly through
their local libraries, or via inter-library loan agreements

[BioMed Central:] This is not surprising, since the Joint Information
Systems Committee (JISC) recently signed a 'big deal', giving access to
Elsevier content for UK academic researchers. But rather missing from
Elsevier's submission is the fact that, for example, the UK National
Health Service (NHS) does not have a national Science Direct deal, so most
clinicians and medical researchers within the NHS, unaffiliated with
academic institutions, do not have access to Elsevier content. Nor do many
small biotech companies - in fact, the difficulty and cost of gaining
access to the latest research is a significant issue for them. And
obviously, the man in the street has essentially no options for online
access.

[Elsevier:] * Per article costs for customers: In the case of Elsevier,
the average cost for a retrieved article for UK users of ScienceDirect has
fallen from �4.57 to �1.69 since 2001, a reduction of 63%. We estimate the
cost to customers per article downloaded will be less than �1 within two
years.

[BioMed Central:] Due to the use of more efficient systems, designed from
the ground up around online Open Access publishing, publishing houses such
as BioMed Central can already sustainably beat �1 per access (in total
cost to the scientific community), even though we have had to start from
scratch and don't yet have the full potential of economies of scale
available to a very large publisher like Elsevier. As the scale of open
access increases, the differential will grow.

[Elsevier:] Open Access' author-pays model risks penalising the UK because
British researchers produce a disproportionately high number of articles
every year. By charging authors for each article that has been accepted
for publication, Open Access transfers the costs of publishing from
institutions like commercial corporations, and libraries that serve
readers, to researchers and their sponsors (e.g. universities,
governmental funding agencies and foundations). While Britain's spending
on journal subscriptions currently amounts to 3.3% of the world's total,
UK researchers contribute a much higher 5% of all articles published
globally. As a result, we estimate that the UK Government, foundations,
universities and researchers could together pay 30-50% more for STM
journals in an Open Access system than they do today.

Whilst individual institutions like Cambridge University and Imperial
College London that are relatively prolific would pay more under an Open
Access system, by contrast, commercial organisations that subscribe to
many journals but contribute relatively few articles each year would pay
substantially less: our estimates suggest that some commercial
corporations would pay one tenth or less in an Open Access system than
they pay under today's subscription model.

[BioMed Central:] Scientists and institutions benefit from making their
published research available to a wide audience - it is by publishing
influential research that institutions acquire a reputation that brings
them high levels of funding and top researcher. And the cost of
dissemination is tiny compared to the cost of doing the research in the
first place.  So it is perfectly reasonable for a top institution to pay
for the publication process that ensures that it's research is effectively
disseminated. And as it happens, most of the costs of scientific
publishing are on the manuscript processing/peer review side of things
anyway - so this form of payment is actually the most direct. Whereas
providing online access to published content is extremely cheap, in an
online environment.  In the traditional environment, the less well off
institutions, which publish little research, effectively subsidize
(through subscriptions) the publication costs better off institutions,
which publish a lot.

Page 2.

[Elsevier:] In addition to these cost-transfer effects, there are other
key unresolved issues concerning Open Access:  

* By introducing an author-pays model, Open Access risks undermining
public trust in the integrity and quality of scientific publications that
has been established over hundreds of years. The subscription model, in
which the users pay (and institutions like libraries that serve them),
ensures high quality, independent peer review and prevents commercial
interests from influencing decisions to publish. This critical control
measure would be removed in a system where the author-or indeed his/her
sponsoring institution-pays. Because the number of articles published will
drive revenues, Open Access publishers will continually be under pressure
to increase output, potentially at the expense of quality.

[BioMed Central:] If a student pays tuition fees, does that make his exam
easier to pass? The overwhelming majority of the journals published by
Elsevier have traditionally seen price increases proportional with the
increase in their volume (plus inflationary increases and increases to
compensate for subscription attrition). As a result, they would benefit
from a higher acceptance rate in the same way that they imply Open Access
publishers do. And they will have to resist the same temptation to lower
the acceptance thresholds in the interest of maintaining the journals'
reputation.

[Elsevier:] * The Open Access business model in its current form has not
proven its financial viability:  even the highest article fees charged by
Open Access publishers today ($1,500) cover only about 40% - 60% of the
estimated total costs to publish an article of the quality that
researchers are used to today.

[BioMed Central:] Elsevier's cost estimates are most likely based on the
inefficient operations of traditional publishers, who have bolted online
offerings onto print, rather than focusing on the efficiencies of a system
designed from first principles for efficient online manuscript submission,
peer review, and publication.

[Elsevier:] Remaining costs, estimated to range from �1 billion - �2
billion for the industry globally, would have to be covered by foundation,
university and government subsidies. While it is conceivable that mean
costs per article may fall as electronic-only publishers gain scale
(currently less than 1% of articles are Open Access), Open Access
publishers are unlikely to cover production costs with revenues of just
$1,500 per article, assuming they provide similar levels of quality, peer
review, functionality and accessibility as researchers receive today. They
would almost certainly be unable to invest in technological innovation to
any significant extent or in nurturing emerging areas of science.

[BioMed Central:] One of the Elsevier's recurring themes seems to be: "we
need to make very high profits to justify investment/innovation so that we
(the largest player in the industry), can, almost like a benevolent
dictator, deliver this innovation to the grateful populace."  The key
response to this is that it is not huge investment by a large corporation
that best drives innovation in the online world. Open platforms drive
innovation, as the internet has shown. Because anyone can plug a site into
the internet and offer any imaginable online service, it has produced
innovation like nothing seen before. The open scientific literature offers
the same kind of promise for science. Once the majority of the scientific
literature is Open Access (in the full sense, of being openly
redistributable), the sky is the limit as far as innovation is concerned,
as the entire scientific community is free to develop and improve the
technology that is being used. We can only imagine the barest outline of
what is possible, but what is certain is that it will dwarf whatever
Elsevier might achieve with their mega investment.  [As a cautionary tale
of how throwing money at a problem isn't the answer, RedSage was a vastly
expensive centralized digital journal project undertaken by AT&T and
various large publishers, which was rapidly superceded and made irrelevant
by smaller scale full text journal website projects such as the Journal of
Biological Chemistry.  http://www.springer-ny.com/press/redsage/

[Elsevier:] * For universal access to be a reality, publishers must
continue to make articles available in multiple media formats. Print is
used by many scientists around the world and by global citizens who are
the beneficiaries of scientific and medical research. To rely on the
internet alone for distribution, as most Open Access journals do, risks
reducing levels of access among these beneficiaries: only 11% of the
world's population uses the Internet and only 64% of UK citizens have ever
been online.

[BioMed Central:] It is somewhat bizarre that Elsevier imagines that the
89% of the world's population who have never used the internet are somehow
likely to have access to print copies of Elsevier journals.  Meanwhile,
the 11% figure argues that there are 600 million people who are only a
click away from an open access article.  In any case, it is odd for them
to say that Open Access journals don't offer print. Both of the most
prominent Open Access publishers (BioMed Central and PLoS) have extensive
print offerings, and the logic that some people may want to pay for print
has very little bearing on open access. But worth noting that:
 (a) Many of the libraries who receive a print copy of BioMed Central
journals that have a subscription component, such as Genome Biology, have
asked us not to send the print, as they actively find print a problem.
 (b) We offer any library the opportunity to receive, at cost, a print
archival copy of all or a portion of the research that we publish. Not one
library has so far taken us up on the offer. Print seems to be of rapidly
decreasing importance to libraries.

[Elsevier:] The recent period of rapid, intense innovation in STM
publishing-the context in which Open Access has emerged-is far from over.
As this period continues, we expect the measurable benefits in
productivity for users (i.e. access, usage, functionality and lower unit
costs for customers) to continue. Elsevier, like all publishers, will
continue to innovate, to observe the impact of innovations like Open
Access and to assess how effectively such initiatives serve the needs of
scientific and research communities. As developments prove able to bring
demonstrable, substantial and sustainable improvements for those
communities, Elsevier will adapt and invest accordingly. In the meantime,
we believe that the UK Government should continue to allow the market
dynamics of this global industry to drive innovation and to determine
which publishing models can best serve the needs of the worldwide
scientific and medical research communities.

Page 6. 

A competitive marketplace 

STM markets are highly competitive: among the 2,000 STM publishers none
has disproportionate power.

[BioMed Central:] Depending on what 'disproportionate' is taken to mean,
of course.

[Elsevier:] The UK Office of Fair Trading (OFT) noted in September 2002
"the overall market is fragmented, with the top six publishers accounting
for just 37% of rated journals and 44% of articles. The top publisher
(Elsevier) accounted for just 13% of the journals, rising to 18% following
its merger with Harcourt."

Frequently, publishers launch new journals as new scientific disciplines
emerge, and as research output grows. (For most of the 20th century, the
number of STM journals has grown, on average, at a reasonably steady 3.25%
per year.) New publishers, sometimes experimenting with new business
models, have also entered the market recently, reflecting the low barriers
to entry.

[BioMed Central:] The number of journals is not a very meaningful metric.
The variation in size (number of articles published in a year) is vast. In
the Elsevier journal programme they range from quarterlies, with perhaps
some 30-40 articles a year, to journals that come out with
telephone-directory-sized issues every week, containing thousands of
articles a year.

[Elsevier:] A dynamic market rapidly changing due to technology and the
progression of science As technological innovation and the advancement of
science continue to change the market, STM publishers respond in a variety
of ways: by launching titles as new sub-disciplines arise, changing the
scope of existing titles, making sure that the journals' editors fairly
represent the leading edge in their fields, and adding search, retrieval
and display functionality to improve the productivity of research. As more
articles become available electronically, usage statistics have enabled
library customers, who have always had discretion about how to spend their
budgets, to make even more informed spending choices.

[BioMed Central:] Selling subscription access to original scientific
research is unlike a typical competitive market, because scientific
publications are not substitutable. If research is published in a
particular journal, that is the only place you can get it. So once a
journal has become established and authors submit to it, a publisher has
an undue degree of leverage when selling that research back to the
scientific community. Essentially the research becomes hostage to the
publisher.

Long term, of course, a publisher cannot exploit its position
indefinitely. Ultimately the institutions who are being asked for
excessive payment will take the difficult decision to walk away, and to
cope as best they can without access to the content. And gradually authors
too will publish elsewhere.  There is evidence that this is starting to
happen. There has been a wave of cancellations of 'big deals' in the last
few months, as they came up for renewal. But scientific institutions and
funders must be aware that they can unwittingly act as obstacles to change
- for example, our research shows that many scientists would very much
like to publish their research in an open access journal, but are held
back by the perception that their funder or review body will not regard a
publication in a new journal in as favourable a light as a publication in
a well established traditional journal.  The funding agencies/review
bodies are therefore not bystanders - they are active participants in
determining the way the scientific publication process works. Therefore it
is not an option for those funding agencies/review bodies to 'leave things
up to the market'. They are already deeply involved in shaping the current
scientific publication marketplace, and tilting it in favour of the
established traditional publishers, who are understandably intent on
defending their historically highly profitable industry. Meanwhile,
technology has made possible an alternative to the traditional model of
publishing. This Open Access model far better serves the needs of
scientist, but it is new and relatively unknown. The funders and review
bodies must decide whether to defend the interests of science and the
scientific community by leveling the playing field for new Open Access
journal publishers, or instead, to (wittingly or unwittingly) defend the
stranglehold that traditional publishers have on the control of the
scientific literature.

The single most important thing that funders and review bodies can do is
to ensure that research published in Open Access journals is judged on
it's own merits, an it not considered less worthy than the same research
published in a traditional journal. And to make it absolutely,
unambiguously publicly clear that this is their position.  A number of the
most prestigious funding bodies are leading the way and are already doing
just that.

Page 8.

[Elsevier:] Key points:

* Open Access' authors'-pay-per-article model risks penalising the UK
because British researchers produce a disproportionately high number of
articles every year. While Britain's spending on journal subscriptions
currently amounts to 3.3% of the world's total, UK researchers contribute
a much higher 5% of all articles published globally. One consequence of
increasing numbers of Open Access journals is that UK researchers and
their sponsors could together pay 30-50% more for STM journals than they
do today.

* In an Open Access system, costs will be transferred to relatively
prolific nations (like the UK) and institutions, like Cambridge University
and Imperial College London, who will pay more. Less prolific
institutions, particularly commercial corporations, will pay much less -
in some cases as little as one tenth of what they pay today.

* Open Access in its current form has not proven its financial viability:
author fees cover only 40%-60% of the estimated costs to publish articles
at the levels of quality that researchers are used to, with remaining
costs, (estimated to range from �1billion - �2 billion for the industry
globally) currently covered by university, foundation and government
agency subsidies.

* While mean costs per article could fall as electronic-only publishers
gain scale (currently less than 1% of articles are Open Access), we
estimate that they would have to fall by as much as 40% - 60% for the
British academic system to pay the same in an Open Access system as it
does today.xiii Reductions of this magnitude would almost certainly mean
that publishers would have difficulty maintaining today's high quality of
STM journals, and would have little if any margin to continue investing in
technology and in nurturing emerging areas of science.

[BioMed Central:] This whole thing boils down to 'we need to be allowed to
squeeze inflated monopoly-driven profits out of the market, because only
inflated monopoly profits give us the resources to do good things."  This
reasoning is fundamentally deeply flawed, but to address a couple of
specific points:

Re: investing in technology:  The strongest stimulus for technology is an
open platform such as the internet.  Nothing would drive technological
innovation in science publishing as much as an open body of scientific
content. E.g. the entire e-science program could achieve radically more if
full-text research articles could be accessed and mined freely through the
'grid' infrastructure.  Not to mention that Open Access publishers have a
strong incentive to invest in technology to reduce their costs, since they
are operating in a fully competitive market for provision of publishing
services (compared to the very imperfect competition between publishers of
subscription research journals)

Re; nurturing emerging areas of science: 

An Open Access publishing environment means that it is no longer necessary
for there to be a commercially viable market for subscriptions to a
journal, in order for that journal to exist (there is no minimum number of
subscriptions below which the journal is not viable, or a market that can
sustain high prices that come with low numbers of subscriptions). This
allows journals to develop in new niches that would have been too small or
too poor to support a traditional journal. BioMed Central has published
several journals that show how the previous publishing models had failed
to cover a particular are -e.g. Malaria Journal.

Page 9.

[Elsevier:] The current state of Open Access 

Open Access is in its infancy, representing less than 1% of published STM
articles. Its journals typically offer only basic text and images with
virtually no or limited search and cross-linking functionality.

[BioMed Central:] Not in the least true with regard to basic text,
searching and linking - see below for a more detailed discussion.

[Elsevier:] Unlike traditional STM publications that may be distributed
via print and online, Open Access journals are typically distributed via
the Internet only. This limits the availability of Open Access journals to
those researchers in nations and institutions that have the required
technological infrastructure. It also limits general availability: only
64% of UK adults have ever used the Internet

[BioMed Central:] Firstly, again this is wholly incorrect - there is
nothing about the open access model which prevents a print copy being
available to subscribers who want it, and pretty much all open access
publishers, including PLoS and BMC, offer this.

Secondly: it is 68% now! How many UK adults have ever gone into a
scientific reference library? The fact that Open Access makes the
scientific literature literally a click away from 68% (and rising) of UK
adults (including journalists, policy makers, MPs to name just a few) is a
huge argument for Open Access.

[Elsevier:] Implications of increasing numbers of Open Access journals

[BioMed Central:] This whole section on how much it costs to publish an
article is largely moot. It can't cost the global scientific community any
more to publish open access than it does to publish in subscription
journals and then pay publishers for their costs, with the generous profit
margins on top. It seems likely that Open Access publishers can reduce the
costs vastly compared to traditional publishers and their inefficient
systems, and so the UK will end up paying less, in an Open Access
publishing environment, than it does now.  But in any case, it seems
bizarre to argue that subscription payments from countries that publish
little research should subsidize the publication of research in countries
like the UK (that publish a lot). Publishing research costs money (though
it costs vastly less than doing the research in the first place).  
Countries with active research communities (like the UK) will reap huge
rewards from Open Access in terms of the benefits for science, which
should easily justify the costs incurred in supporting the infrastructure
for open access publication.

Page 10.

[Elsevier:] A third factor is that if high quality journals are to remain
in business long term without subsidies, they will likely have to raise
their per-article fees substantially to cover their technology and
editorial costs. If Open Access publishers are to improve the basic
functionality of their current offerings (e.g. by adding search
functionalities, linking and profiling), costs will again increase.

[BioMed Central:] This is a completely baseless comment provided with no
evidence. 

BioMed Central provides an environment for scientists (as peer reviewers,
authors, readers, users) that is at least as technologically rich as
anything Elsevier provides. E.g. advanced search, stored search histories,
customized email alerting, RSS feeds, OpenURL links, Crossref links,
PubMed links, cited-by links, comment on the article in question, instant
publication on acceptance, Open Archives Initiative protocol support, to
name just a few.  We have also developed advanced XML/MathML authoring
tools in collaboration with Wolfram Research, and automated reference
extraction technologies for processing incoming manuscripts in
collaboration with ISI Researchsoft.

[Elsevier:] Alternatively, Open Access publishers may abandon their
no-subscription-fee approach and adopt hybrid models that incorporate a
subscription component, as BioMed Central has already done.

[BioMed Central:] This is puzzling and shows ignorance about what BioMed
Central offers. BioMed Central's model has been clear and consistent from
the moment it started - Open Access to research, subscription access to
some added value online content (e.g. commissioned pieces, editorially
intensive databases etc), and to print editions, where those are desired.

[Elsevier:] Fourth, the quality of research articles might well suffer as
Open Access publishers compromise the rigour of their peer review
processes by rejecting fewer articles so that they can publish more and
increase revenues. Weaker articles, or articles that serve commercial
interests, may therefore get published when previously they would have
been rejected. By contrast, the current subscription system supports a
highly independent peer review process in which publishers actively manage
editorial boards. In an Open Access environment, there could be pressures
from institutions to shape editorial direction as well as on the volume of
submissions.

[BioMed Central:]

(a)  Authors choice what journals to publish in is largely based on the
reputation of that journal, so the journal has every incentive not to
damage its own reputation

(b)  Many highly regarded journals charge page charges for color figures
etc, with no problems for their reputation (e.g. Genes & Development, PNAS
to name just two : www.pnas.org/misc/iforc.shtml ,
http://www.genesdev.org/misc/ifora.shtml )

(c)  The overwhelming majority of the journals published by Elsevier have
traditionally seen price increases proportional with the increase in their
volume (plus inflationary increases and increases to compensate for
subscription attrition). As a result, they would benefit from a higher
acceptance rate in the same way that Open Access publishers do. And they
will have to resist the same temptation to lower the acceptance thresholds
in the interest of maintaining the journals' reputation.

[Elsevier:] Finally, assuming that cited research continues to be the
primary measurement of research institutions' quality, with quality the
basis for funding allocation, it is unclear whether Open Access journals
will affect the Research Assessment Exercise (RAE). ISI, the industry
standard that provides key data for the RAE on the quality of research,
currently measures only two out of some 500 Open Access journals because
the rest are too new or too irregularly published to give valid data.

Presumably this number will increase as Open Access journals become
established but at this point it is too early to assess what portion of
Open Access journals will appear in ISI-rated journals even two years from
now (the minimum time required before ISI will consider rating a journal).
The greater the proportion that meet ISI's strict criteria to be rated,
the less will be the impact on the RAE.

[BioMed Central:] A vast number of issues to raise with the above.

1. Re the comment: "...assuming that cited research continues to be the
primary measurement of research institutions' quality,"  It is a
speculative assumption that citations (whether of the article itself, or
the journal impact factor), is the primary measure of quality. It is well
known that ISI citation figures are radically imperfect as measures of
quality - the omission of new journals being only the most glaring of many
distortions that they produce. Elsevier directors Amin and Mabe* have
written a critique of impact factors and conclude that "...they are not a
direct measure of quality and must be used with considerable care."  Our
understanding is that the RAE explicitly does not base it's reports on
crude impact factor measurements.

2. Re the comment: "ISI, the industry standard that provides key data for
the RAE on the quality of research, currently measures only two out of
some 500 Open Access journals because the rest are too new or too
irregularly published to give valid data."  Again, not true. BioMed
Central alone has 6 journals that currently have impact factors, listed
here:  http://www.biomedcentral.com/info/about/faq?name=impactfactor

ISI explicitly tracks 22 BioMed Central journals and several more of these
will get impact factors in June 2004.  And citations of the other 80+
BioMed Central journals are already captured and tracked in ISI's cited
reference database, so although ISI does not yet produce journal impact
factors for these journals, if one wants to find out how many times an
article has been cited, one can do so.  So it is only the crudest possible
measure of research quality (the journal impact factor) on which this has
any bearing. And surely if ISI journal impact factors fail to adequately
measure the quality of new Open Access journals, this is yet another
reason to limit their use in the RAE. Elsevier's suggestion that the UK
scientific community should avoid Open Access journals to avoid affecting
the RAE is putting the cart totally before the horse.

* Elsevier's earlier opinion on Impact Factors:

M. Amin & M. Mabe, Elsevier Science

IMPACT FACTORS: USE AND ABUSE 

Perspectives in Publishing, Number 1, October 2000 

http://www1.elsevier.com/homepage/about/ita/editors/perspectives1.pdf 

Conclusions (page 6 of the above article) 

"This pamphlet has shown that impact factors are only one of a number of
measures for describing the "impact" that particular journals can have in
the research literature. The value of the impact factor is affected by the
subject area, type and size of a journal, and the "window of measurement"
used. As statistical measures they fluctuate from year to year, so that
great care needs to be taken in interpreting whether a journal has really
"dropped (or risen)" in quality from changes in its impact factor. Use of
the absolute values of impact factors, outside of the context of other
journals within the same subject area, is virtually meaningless; journals
ranked top in one field may be bottom in another. Extending the use of the
journal impact factor from the journal to the authors of papers in the
journal is highly suspect; the error margins can become so high as to make
any value meaningless. Professional journal types (such as those in
medicine) frequently contain many more types of source item than the
standard research journal. Errors can arise in ensuring the right types of
article are counted in calculating the impact factor.

Citation measures, facilitated by the richness of ISI's citation
databases, can provide very useful insights into scholarly research and
its communication. Impact factors, as one citation measure, are useful in
establishing the influence journals have within the literature of a
discipline. Nevertheless, they are not a direct measure of quality and
must be used with considerable care. "

______________________________________________________________________
This email has been scanned by the MessageLabs Email Security System.
For more information please visit http://www.messagelabs.com/email 
______________________________________________________________________